Among the other mountain-ranges may be mentioned the Altai Mountains, in the south-west quadrant (Plate XVI.), the Carpathians, close to the great crater Copernicus, and the beautiful semicircle of hills which borders the Sinus Iridum, or Bay of Rainbows, to the east of the Alpine range. This bay forms one of the loveliest of lunar landscapes, and under certain conditions of illumination its eastern cape, the Heraclides Promontory, presents a curious resemblance, which I have only seen once or twice, to the head of a girl with long floating hair--'the moon-maiden.' The Leibnitz and Doerfel Mountains, with other ranges whose summits appear on the edge of the moon, are seldom to be seen to great advantage, though they are sometimes very noticeably projected upon the bright disc of the sun during the progress of an eclipse.[*] They embrace some of the loftiest lunar peaks reaching 26,000 feet in one of or two instances, according to Schröter and Mädler. [Illustration: FIG. 23. ARISTARCHUS and HERODOTUS, February 20, 1891, 6.15 p.m., 3-7/8 inch.] 4. THE CLEFTS OR RILLS.--In these, and in the ray-systems, we again meet with features to which a terrestrial parallel is absolutely lacking. Schröter of Lilienthal was the first observer to detect the existence of these strange chasms, and since his time the number known has been constantly increasing, till at present it runs to upwards of a thousand. These objects range from comparatively coarse features, such as the Herodotus Valley (Fig. 23), and the well-known Ariadæus and Hyginus clefts, down to the most delicate threads, only to be seen under very favourable conditions, and taxing the powers of the finest instruments. They present all the appearance of cracks in a shrinking surface, and this is the explanation of their existence which at present seems to find most favour. In some cases, such as that of the great Sirsalis cleft, they extend to a length of 300 miles; their breadth varies from half a mile, or less, to two miles; their depth is very variously estimated, Nasmyth putting it at ten miles, while Elger only allows 100 to 400 yards. In a number of instances they appear either to originate from a small crater, or to pass through one or more craters in their course. The student will quickly find out for himself that they frequently affect the neighbourhood of one or other of the mountain ranges (as, for example, under the eastern face of the Apennines, Plate XVII.), or of some great crater, such as Archimedes. They are also frequently found traversing the floor of a great walled-plain, and at least forty have been detected in the interior of Gassendi (Plate XIX., No. 90). Smaller instruments are, of course, incompetent to reveal more than a few of the larger and coarser of these strange features. The Serpentine Valley of Herodotus, the cleft crossing the floor of Petavius, and the Ariadæus and Hyginus rills are among the most conspicuous, and may all be seen with a 2-1/2-inch